Table 1.
Group | Method | Dataset size | Accuracy | Segmentation time |
---|---|---|---|---|
Our group17 | Local appearance and shape model (semiautomatic) | 42 (test and training) | Whole gland:MAD: DSC: Recall: Precision: : | Operator interaction: . (across 10 images and 9 operators) Execution: . (across 42 images, one operator) |
Cheng et al.21 | Atlas-based (automatic) | 100 (training) and 40 (test) | Whole gland:TP: 91.2%DSC: 87.6%: 8.4% | NA |
Liao et al.18 | Multi-atlas-based (automatic) | 66 (test) 9 (atlas) | Whole gland:MAD: DSC: | Execution: 2.9 min |
Toth and Madabhushi19 | Active appearance model (semiautomatic) | 108 | Whole gland:MAD: DSC: | Execution: 150 s |
Vikal et al.22 | Shape model (semiautomatic) | 3 | Has not reported for whole gland | Execution: 23 s |
Martin et al.20 | Atlas-based (semiautomatic) | 1 (reference) 17 (test) | Whole gland:MAD: Recall: Precision: | NA |
Note: MAD, mean absolute distance; DSC, Dice similarity coefficient; , volume difference; TP, true positive.